Multi-Feature Optimization Study of Soil Total Nitrogen Content Detection Based on Thermal Cracking and Artificial Olfactory System

To improve the accuracy of detecting soil total nitrogen (STN) content by an artificial olfactory system, this paper proposes a multi-feature optimization method for soil total nitrogen content based on an artificial olfactory system. Ten different metal–oxide semiconductor gas sensors were selected to form a sensor array to collect soil gas and generate response curves. Additionally, six features such as the response area, maximum value, average differential coefficient, standard deviation value, average value, and 15th-second transient value of each sensor response curve were extracted to construct an artificial olfactory feature space (10×6). Moreover, the relationship between feature space and soil total nitrogen content was used to establish backpropagation neural network (BPNN), extreme learning machine (ELM), and partial least squares regression (PLSR) models were used, and the coefficient of determination (R2), root mean square error (RMSE), and the ratio of performance to deviation (RPD) were selected as prediction performance indicators. The Monte Carlo cross-validation (MCCV) and K-means improved leave-one-out cross-validation (K-means LOOCV) were adopted to identify and remove abnormal samples in the feature space and establish the BPNN model, respectively. There were significant improvements before and after comparing the two rejection methods, among which the MCCV rejection method was superior, where values for R2, RMSE, and RPD were 0.75671, 0.33517, and 1.7938, respectively. After removing the abnormal samples, the soil samples were then subjected to feature-optimized dimensionality reduction using principal component analysis (PCA) and genetic algorithm-based optimization backpropagation neural network (GA-BP). The test results showed that after feature optimization the model indicators performed better than those of the unoptimized model, and the PLSR model with GA-BP for feature optimization had the best prediction effect, with an R2 value of 0.93848, RPD value of 3.5666, and RMSE value of 0.16857 in the test set. R2 and RPD values improved by 14.01% and 50.60%, respectively, compared with those before optimization, and RMSE value decreased by 45.16%, which effectively improved the accuracy of the artificial olfactory system in detecting soil total nitrogen content and could achieve more accurate quantitative prediction of soil total nitrogen content.

Estimating the Spatial Distribution of Soil Organic Carbon and Soil Total Nitrogen Stocks by Remote Sensing Methods: A Case Study in a Forest-dominated Alpine Watershed, Western China

Soil organic carbon and soil total nitrogen stocks are important indicators for evaluating soil health and stability. Accurately predicting the spatial distribution of soil organic carbon and total nitrogen stocks is an important basis for mitigating global warming, ensuring regional food security, and maintaining the sustainable development of ecologically fragile areas. On the basis of field sampling data and remote sensing technology, this study divided the topsoil (0–30 cm) into three soil layers of 0–10 cm, 10–20 cm, and 20–30 cm to carry out soil organic carbon and soil total nitrogen stocks estimation experiments in the Qilian Mountains in western China. A multiple linear regression model and a stepwise multiple linear regression model were used to estimate soil organic carbon and soil total nitrogen stocks. A total of 119 topsoil samples and nine remotely sensed environmental variables were collected and used for model development and validation. The results indicated that these two linear regression models showed good performance. The modified soil-adjusted vegetation index (MSAVI), perpendicular vegetation index (PVI), aspect, elevation, and solar radiation were the key environmental variables affecting soil organic carbon and total nitrogen stocks. In topsoil, remote sensing technology could be used to predict the soil properties in layers; however, as the soil depth increased, the performance of the linear regression models gradually decreased.